CHICAGO – Renal denervation, relative to a sham procedure, was linked with statistically significant reductions in blood pressure in the newly completed SPYRAL HTN–ON MED trial, but several factors are likely to have worked in concert to prevent the study from meeting its primary endpoint.
Of these differences, probably none was more important than the substantially higher proportion of patients in the sham group that received additional BP-lowering medications over the course of the study, David E. Kandzari, MD, reported at the American Heart Association scientific sessions.
The SPYRAL HTN–ON MED pivotal trial followed the previously completed SPYRAL HTN–ON MED pilot study, which did show a significant BP-lowering effect on antihypertensive medications followed radiofrequency denervation. In a recent update of the pilot study, the effect was persistent out to 3 years.
In the SPYRAL HTN–ON MED program, patients on their second screening visit were required to have a systolic pressure of between 140 and 170 mm Hg on 24-hour ambulatory BP monitoring (ABPM) while taking up to three antihypertensive medications. Patients who entered the study were randomized to renal denervation or sham control while maintaining their baseline antihypertensive therapies.
The previously reported pilot study comprised 80 patients. The expansion pivotal trial added 257 more patients for a total cohort of 337 patients. The primary efficacy endpoint was based on a Bayesian analysis of change in 24-hour systolic ABPM at 6 months for those in the experimental arm versus those on medications alone. Participants from both the pilot and pivotal trials were included.
The prespecified definition of success for renal denervation was a 97.5% threshold for probability of superiority on the basis of this Bayesian analysis. However, the Bayesian analysis was distorted by differences in the pilot and expansion cohorts, which complicated the superiority calculation. As a result, the analysis only yielded a 51% probability of superiority, a level substantially below the predefined threshold.
Despite differences seen in BP control in favor of renal denervation, several factors were identified that likely contributed to the missed primary endpoint. One stood out.
“Significant differences in medication prescriptions were disproportionate in favor of the sham group,” reported Dr. Kandzari, chief of Piedmont Heart Institute, Atlanta. He said these differences, which were a violation of the protocol mandate, led to a “bias toward the null” for the primary outcome.
The failure to meet the primary outcome was particularly disappointing in the wake of the favorable pilot study and the SPYRAL HTN–OFF MED pivotal trial, which were both positive.
In the pilot study, which did not have a medication imbalance, a 7.3–mm Hg reduction (P = .004) in 24-hour ABPM was seen at 6 months. Relative reductions in office-based systolic pressure reductions for renal denervation versus sham were 6.6 mm Hg (P = .03) and 4.0 mm Hg (P = .03) for the pilot and expansions groups, respectively.
On the basis of a Win ratio derived from a hierarchical analysis of ABMP and medication burden reduction, the 1.50 advantage (P = .005) for the renal denervation arm in the newly completed SPYRAL HTN–ON MED trial was also compelling.
At study entry, the median number of medications was 1.9 in both the renal denervation and sham arms. At the end of 6 months, the median number of medications was unchanged in the experimental arm but rose to 2.1 (P = .01) in the sham group. Similarly, there was little change in the medication burden from the start to the end of the trial in the denervation group (2.8 vs. 3.0), but a statistically significant change in the sham group (2.9 vs. 3.5; P = .04).
Furthermore, the net percentage change of patients receiving medications favoring BP reduction over the course of the study did not differ between the experimental and control arms of the pilot cohort, but was more than 10 times higher among controls in the expansion group (1.9% vs. 21.8%; P < .0001).
Medication changes over the course of the SPYRAL HTN–ON MED trial were even greater in some specific subgroups. Among Black participants, for example, 14.2% of those randomized to renal denervation and 54.6% of those randomized to the sham group increased their antihypertensive therapies over the course of the study.
The COVID-19 epidemic is suspected of playing another role in the negative results, according to Dr. Kandzari. After a brief pause in enrollment, the SPYRAL HTN–ON MED trial was resumed, but approximately 80% of the expansion cohort data were collected during this period. When compared, variances in office and 24-hour ABPM were observed for participants who were or were not evaluated during COVID.
“Significant differences in 24-hour ABPM patterns pre- and during COVID may reflect changes in patient behavior and lifestyle,” Dr. Kandzari speculated.
The data from this study differ from essentially all of the other studies in the SPYRAL HTN program as well as several other sham-controlled studies with renal denervation, according to Dr. Kandzari.
The AHA-invited discussant, Ajay J. Kirtane, MD, director of the Cardiac Catheterization Laboratories at Columbia University, New York, largely agreed that several variables appeared to conspire against a positive result in this trial, but he zeroed in on the imbalance of antihypertensive medications.
“Any trial that attempts to show a difference between renal denervation and a sham procedure must insure that antihypertensive medications are the same in the two arms. They cannot be different,” he said.
As an active investigator in the field of renal denervation, Dr. Kirtane thinks the evidence does support a benefit from renal denervation, but he believes data are still needed to determine which patients are candidates.
“Renal denervation is not going to be a replacement for previous established therapies, but it will be an adjunct,” he predicted. The preponderance of evidence supports clinically meaningful reductions in BP with this approach, “but we need to determine who to consider [for this therapy] and to have realistic expectations about the degree of benefit.”
Dr. Kandzari reported financial relationships with Abbott Vascular, Ablative Solutions, Biotronik, Boston Scientific, CSI, Medtronic Cardiovascular, OrbusNeich, and Teleflex. Dr. Kirtane reported financial relationships with Abbott Vascular, Abiomed, Boston Scientific, Cardiovascular Systems, Cathworks, Chiesi, Medtronic, Opens, Philipps, Regeneron, ReCor Medical, Siemens, Spectranetics, and Zoll.